Model A

Xiaomi

65.41/100

Supported · Public rank #41

90% interval 56.8–74.0

MiMo-V2.5-Pro vs Qwen3.6-35B-A3B

Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.

Model B

Qwen3.6-35B-A3B

Alibaba

48.34/100

Estimated · Public rank #150

90% interval 42.6–54.1

Decision reading

MiMo-V2.5-Pro has the higher public score, 65.41 versus 48.34, and the 90% score intervals do not overlap.

7 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.

Share or export

Share on X LinkedInSocial card CSV JSON

Which one for your work

Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.

Long documents
Prompts that approach the documented context limit
MiMo-V2.5-Pro
MiMo-V2.5-Pro has the larger documented context window.
Confidence: documented

Show secondary and unsupported calls

Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
MiMo-V2.5-Pro and Qwen3.6-35B-A3B are scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Qwen3.6-35B-A3B is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
Confidence: limited
Chat turn cost
1K fresh input + 500 output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
Not enough matched evidence
A complete comparable API-rate estimate is not available for both models.
Confidence: listed-rates

What is actually comparable

Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.

6 MiMo-V2.5-Pro only

7 Shared

34 Qwen3.6-35B-A3B only

Shared results: 7
MiMo-V2.5-Pro only: 6
Qwen3.6-35B-A3B only: 34
Like-for-like categories: 0 / 8

4 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.

Category results, on a stated basis

Each row shows the public-lane category score for both models: the BenchAlign lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.

Agentic

Directional only

MiMo-V2.5-Pro: 40.8; Supported · #118/151
Qwen3.6-35B-A3B: 44.4; Estimated · #102/151
Basis: BenchAlign lane · 5 vs 11 public rows
Reading: Directional only

Coding

Directional only

MiMo-V2.5-Pro: 57.1; Estimated · #36/183
Qwen3.6-35B-A3B: 47.3; Estimated · #90/183
Basis: BenchAlign lane · 4 vs 6 public rows
Reading: Directional only

Knowledge

Directional only

MiMo-V2.5-Pro: 55.5; Supported · #58/181
Qwen3.6-35B-A3B: 46.5; Estimated · #109/181
Basis: BenchAlign lane · 4 vs 5 public rows
Reading: Directional only

Instruction following

Directional only

MiMo-V2.5-Pro: 93.5; #6/120
Qwen3.6-35B-A3B: 78.0; #55/120
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Directional only

Reasoning

Not comparable

MiMo-V2.5-Pro: 76.9; Unranked · 2 rankable rows
Qwen3.6-35B-A3B: 69.0; Unranked · 2 rankable rows
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Math

Not comparable

MiMo-V2.5-Pro: Not ranked
Qwen3.6-35B-A3B: 71.5; Unranked · 5 rankable rows
Basis: Provisional lane · 0 vs 2 weighted rows
Reading: Not comparable

Multilingual

Not comparable

MiMo-V2.5-Pro: Not ranked
Qwen3.6-35B-A3B: Not ranked
Basis: Provisional lane · 0 vs 0 weighted rows
Reading: Not comparable

Multimodal

Not comparable

MiMo-V2.5-Pro: Not ranked
Qwen3.6-35B-A3B: 50.1; #37/48
Basis: Provisional lane · 0 vs 2 weighted rows
Reading: Not comparable

Category scores with the server-provided evidence basis for MiMo-V2.5-Pro and Qwen3.6-35B-A3B
Category	MiMo-V2.5-Pro	Qwen3.6-35B-A3B	Basis	Reading
Agentic	40.8Supported · #118/151	44.4Estimated · #102/151	Directional onlyBenchAlign lane · 5 vs 11 public rows	Directional only
Coding	57.1Estimated · #36/183	47.3Estimated · #90/183	Directional onlyBenchAlign lane · 4 vs 6 public rows	Directional only
Knowledge	55.5Supported · #58/181	46.5Estimated · #109/181	Directional onlyBenchAlign lane · 4 vs 5 public rows	Directional only
Instruction following	93.5#6/120	78.0#55/120	Directional onlyProvisional lane · 0 vs 0 weighted rows	Directional only
Reasoning	76.9Unranked · 2 rankable rows	69.0Unranked · 2 rankable rows	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Math	Not ranked	71.5Unranked · 5 rankable rows	Not comparableProvisional lane · 0 vs 2 weighted rows	Not comparable
Multilingual	Not ranked	Not ranked	Not comparableProvisional lane · 0 vs 0 weighted rows	Not comparable
Multimodal	Not ranked	50.1#37/48	Not comparableProvisional lane · 0 vs 2 weighted rows	Not comparable

Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.

Shape of the matched evidence

Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.

Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.

HLE
Knowledge
MiMo-V2.5-Pro: 48%Qwen3.6-35B-A3B: 21.4%Normalized gap 26.6MiMo-V2.5-Pro source Qwen3.6-35B-A3B source
Terminal-Bench 2.0
Agentic
MiMo-V2.5-Pro: 68.4%Qwen3.6-35B-A3B: 51.5%Normalized gap 16.9MiMo-V2.5-Pro source Qwen3.6-35B-A3B source
SWE-bench Pro
Coding
MiMo-V2.5-Pro: 57.2%Qwen3.6-35B-A3B: 49.5%Normalized gap 7.7MiMo-V2.5-Pro source Qwen3.6-35B-A3B source

What each workload costs

Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.

Chat turn

1K fresh input + 500 output tokens

MiMo-V2.5-Pro: API rate not published; Fits in one request
Qwen3.6-35B-A3B: API rate not published; Fits in one request

MiMo-V2.5-Pro has no comparable published API token rate. Qwen3.6-35B-A3B has no comparable published API token rate.

Repository review

50K fresh input + 3K output tokens

MiMo-V2.5-Pro: API rate not published; Fits in one request
Qwen3.6-35B-A3B: API rate not published; Fits in one request

MiMo-V2.5-Pro has no comparable published API token rate. Qwen3.6-35B-A3B has no comparable published API token rate.

Cache-heavy agent loop

200K cached + 20K fresh input + 10K output tokens

MiMo-V2.5-Pro: API rate not published; Fits in one request; Cached-input rate unavailable
Qwen3.6-35B-A3B: API rate not published; Fits in one request; Cached-input rate unavailable

MiMo-V2.5-Pro has no comparable published API token rate. Qwen3.6-35B-A3B has no comparable published API token rate.

Specification differences

Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.

SpecificationMiMo-V2.5-ProQwen3.6-35B-A3B

Context window

Maximum documented context; output-token limits may be lower.

MiMo-V2.5-Pro

Qwen3.6-35B-A3B

262K

API model ID

MiMo-V2.5-Pro

Not sourced

Qwen3.6-35B-A3B

Not sourced

Cached-input rate

A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.

MiMo-V2.5-Pro

No comparable hosted API rate

Qwen3.6-35B-A3B

No comparable hosted API rate

Documented inputs

MiMo-V2.5-Pro

Not sourced

Qwen3.6-35B-A3B

Not sourced

Documented outputs

MiMo-V2.5-Pro

Not sourced

Qwen3.6-35B-A3B

Not sourced

Provider availability

MiMo-V2.5-Pro

Not sourced

Qwen3.6-35B-A3B

Not sourced

Reasoning profile

MiMo-V2.5-Pro

Reasoning

Qwen3.6-35B-A3B

Reasoning

Weight access

MiMo-V2.5-Pro

Proprietary

Qwen3.6-35B-A3B

Open Weight

License

MiMo-V2.5-Pro

Proprietary

Qwen3.6-35B-A3B

Open Weight

Release date

MiMo-V2.5-Pro

2026-04-22

Qwen3.6-35B-A3B

2026-04-15

If you already use one of these models

Deployment change: The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
Quality signal: MiMo-V2.5-Pro has the higher public score, 65.41 versus 48.34, and the 90% score intervals do not overlap.
Workload cost: A complete comparable API-rate estimate is not available for both models.
Context tradeoff: MiMo-V2.5-Pro has the larger documented window (1M).

Run the same representative tasks against both endpoints before changing production traffic.

Benchmark evidence

The full public result ledger is available for audit without forcing a wide desktop table onto a phone.

Browse raw public benchmark evidence47 rows

Agentic

Claw-Eval
MiMo-V2.5-Pro63.8%
Source
Qwen3.6-35B-A3B68.7%
Source
Qwen3.6-35B-A3B leads this result
τ³-bench results
MiMo-V2.5-Pro72.9%
Source
Qwen3.6-35B-A3B67.2%
Source
MiMo-V2.5-Pro leads this result
Terminal-Bench 2.0
MiMo-V2.5-Pro68.4%
Source
Qwen3.6-35B-A3B51.5%
Source
MiMo-V2.5-Pro leads this result
Gert Labs
Shared source
MiMo-V2.5-Pro62.70%
Qwen3.6-35B-A3B42.65%
MiMo-V2.5-Pro leads this result
Terminal-Bench 2.1 (Vals)
MiMo-V2.5-Pro57.3%
Source
Qwen3.6-35B-A3B—
Not directly comparable
QwenClawBench
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B52.6%
Source
Not directly comparable
QwenWebBench
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B1397
Source
Not directly comparable
VITA-Bench
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B35.6%
Source
Not directly comparable
DeepPlanning
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B25.9%
Source
Not directly comparable
Toolathlon
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B26.9%
Source
Not directly comparable
MCP Atlas
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B62.8%
Source
Not directly comparable
WideResearch
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B60.1%
Source
Not directly comparable

Coding

SWE-bench Pro
MiMo-V2.5-Pro57.2%
Source
Qwen3.6-35B-A3B49.5%
Source
MiMo-V2.5-Pro leads this result
Terminal-Bench 2.0
MiMo-V2.5-Pro68.4%
Source
Qwen3.6-35B-A3B51.5%
Source
MiMo-V2.5-Pro leads this result
LiveCodeBench (Vals)
MiMo-V2.5-Pro81.4%
Source
Qwen3.6-35B-A3B—
Not directly comparable
SWE-bench (Vals)
MiMo-V2.5-Pro74.0%
Source
Qwen3.6-35B-A3B—
Not directly comparable
SWE-bench Verified
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B73.4%
Source
Not directly comparable
SWE Multilingual
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B67.2%
Source
Not directly comparable
LiveCodeBench
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B80.4%
Source
Not directly comparable
NL2Repo
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B29.4%
Source
Not directly comparable

Knowledge

HLE
MiMo-V2.5-Pro48%
Source
Qwen3.6-35B-A3B21.4%
Source
MiMo-V2.5-Pro leads this result
HLE w/o tools
MiMo-V2.5-Pro34%
Source
Qwen3.6-35B-A3B—
Not directly comparable
GPQA Diamond (Vals)
MiMo-V2.5-Pro82.6%
Source
Qwen3.6-35B-A3B—
Not directly comparable
MMLU-Pro (Vals)
MiMo-V2.5-Pro84.6%
Source
Qwen3.6-35B-A3B—
Not directly comparable
MMLU-Pro
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B85.2%
Source
Not directly comparable
SuperGPQA
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B64.7%
Source
Not directly comparable
C-Eval
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B90%
Source
Not directly comparable
GPQA
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B86%
Source
Not directly comparable

Math

HMMT Feb 2025
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B90.7%
Source
Not directly comparable
HMMT Nov 2025
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B89.1%
Source
Not directly comparable
HMMT Feb 2026
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B83.6%
Source
Not directly comparable
MMAnswerBench
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B78.9%
Source
Not directly comparable
AIME26
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B92.7%
Source
Not directly comparable

Multimodal

MMMU
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B81.7%
Source
Not directly comparable
MMMU-Pro
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B75.3%
Source
Not directly comparable
RealWorldQA
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B85.3%
Source
Not directly comparable
OmniDocBench 1.5
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B89.9%
Source
Not directly comparable
CharXiv
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B78%
Source
Not directly comparable
SimpleVQA
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B58.9%
Source
Not directly comparable
CC-OCR
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B81.9%
Source
Not directly comparable
AI2D_TEST
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B92.7%
Source
Not directly comparable
RefCOCO (avg)
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B92.0%
Source
Not directly comparable
ODINW13
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B50.8%
Source
Not directly comparable
Video-MME (with subtitle)
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B86.6%
Source
Not directly comparable
Video-MME (w/o subtitle)
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B82.5%
Source
Not directly comparable
VideoMMMU
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B83.7%
Source
Not directly comparable
MLVU (M-Avg)
MiMo-V2.5-Pro—
Qwen3.6-35B-A3B86.2%
Source
Not directly comparable

Frequently asked questions

Which is better, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

MiMo-V2.5-Pro has the higher public score, 65.41 versus 48.34, and the 90% score intervals do not overlap. That is the clearest overall quality signal in the current public data.

Which is better for coding, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

MiMo-V2.5-Pro scores higher for coding on the public lane, 57.1 to 47.3. MiMo-V2.5-Pro and Qwen3.6-35B-A3B are scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which is better for agentic tasks, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Qwen3.6-35B-A3B scores higher for agentic tasks on the public lane, 44.4 to 40.8. Qwen3.6-35B-A3B is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.

Which costs less, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Both models do not have comparable published API token rates, so this page does not name a universal price winner.

Which has the larger context window, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

MiMo-V2.5-Pro has the larger documented context window: 1M, compared with 262K.

Related comparisons

Compare API pricing Read the methodology Open the model selector

Last updated September 4, 2026

Watch MiMo-V2.5-Pro vs Qwen3.6-35B-A3B

One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.

Read a sample issue

Join 2,000+ readers.

MiMo-V2.5-Pro vs Qwen3.6-35B-A3B

Which one for your work

Long documents

Coding work

Agentic work

Chat turn cost

Cache-heavy agent loop cost

Repository review cost

What is actually comparable

Category results, on a stated basis

Agentic

Coding

Knowledge

Instruction following

Reasoning

Math

Multilingual

Multimodal

Shape of the matched evidence

What each workload costs

Chat turn

Repository review

Cache-heavy agent loop

Specification differences

Context window

API model ID

Cached-input rate

Documented inputs

Documented outputs

Provider availability

Reasoning profile

Weight access

License

Release date

Benchmark evidence

Agentic

Coding

Knowledge

Math

Multimodal

Frequently asked questions

Which is better, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Which is better for coding, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Which is better for agentic tasks, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Which costs less, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Which has the larger context window, MiMo-V2.5-Pro or Qwen3.6-35B-A3B?

Related comparisons

Watch MiMo-V2.5-Pro vs Qwen3.6-35B-A3B